Transitioning from experienced DBA/Developer to Data Engineering — Advice Needed?

[sandeeppri]

Hi Experts,

I have 18 years of experience in database administration, database development, SSRS/Power BI,and data analysis across SQL, Oracle, Postgres, and MySQL.

Looking to transition into Data Engineering, I have chosen to pursue the DP-700 (Microsoft Fabric Data Engineering) certification to guide my upskilling. However, I want to ensure my learning directly aligns with what employers are looking for to secure a job in today's highly saturated market. 

Could you please assist me with the following?

1. Skill Gap: Besides DP-700 and my SQL background, what programming languages (e.g., Python, Spark) or tools are absolute must-haves for Data Engineering roles? 
2. Resume Strategy: How can I best leverage my extensive database administration and development background on my resume to stand out in a competitive job market? 
3. Portfolio Projects: What kind of practical portfolio projects can someone at my level build to demonstrate cloud-native Data Engineering capabilities to hiring managers? 

Any guidance or mentorship you can provide would be greatly appreciated.

Thank you so much,

Sandeep

[AllisClaramunt]

With 18 years of database experience, you already have one of the hardest parts of data engineering (understanding data, performance, modeling, and enterprise constraints...).

The biggest shift is usually moving from managing databases to designing data platforms: ingestion patterns, orchestration, cloud storage, distributed processing, governance, and scalable architectures.

For the Fabric ecosystem, I would focus on:

• Python fundamentals for data engineering workflows
• Spark / PySpark for distributed processing
• Lakehouse concepts (Delta, medallion architecture, OneLake)
• Data pipelines and orchestration patterns
• Security and governance design

Given your SQL and BI background, I would not underestimate semantic modeling and business understanding since many engineering profiles lack that context.

For portfolio projects, I would avoid simple ETL demos and build something closer to an enterprise scenario: ingesting multiple sources, applying transformations, implementing governance/security, and exposing trusted data for analytics or AI use cases.

Your database background is not something to hide on your resume. In fact, it is your differentiator.

Position it as enterprise data platform experience, not only DBA experience.

[GilbertQ]

Hi @sandeeppri 

Just to echo what the other people have said, because you have got an experience in database systems you have done the hard work and understand hard data pieces together as well as with power BR how to model data. So I would recommend that you could always look from a pure ETL perspective in terms of data engineering where you can ingest data and bring it into lake houses or warehouses which is something you can build on with your existing knowledge you. Might have to learn a bit of puffin and understand how delta table works, but that's not too far or too different from what your existing experience has. I would highly recommend just getting started and stuck in and you will then see how similar it actually is.

[BrandonHart]

No two job descriptions for Data Engineers read the same. To be honest it is impossible to know everything. For example, I never used DBT, and most likely most of Fabric/ Databricks Data Engineers probably have not, but some roles absolutely require it. But I would say Python, SQL and cloud based data stores (datalakes, blob storage) . No parquet files, Delta tables and the Pyspark, Request libraries. Then have a great understanding of Medallion architecture. Also undertand Realtime(KQL, Event House) and appropriate use cases for. This will be covered in DP 700 but the concepts will stay the same.

With your experience I think getting the DP 700 and if you can get to the point you feel Dangerous in Python would be enough. Microsoft has made it so that you can do a lot without coding but once you can code, you turn into NEO. There is no spoon.

One last thought. Oddly enough, a lot of the hiring managers for Data Engineers are are not very technical. So it is important to learn how to talk about everything in layman's terms and thoroughly explain concepts in a way that instills confidence that you really know it.

[ssrithar]

Hi  , 

You already have one of the hardest parts that many aspiring data engineers don't—18 years of experience working with data. The challenge isn't starting over; it's demonstrating that you can build modern cloud-native data platforms.

Here are the areas I'd focus on:

1. Build on your existing strengths

Your experience with SQL, Oracle, PostgreSQL, MySQL, reporting, and analytics is highly relevant. Data engineering is fundamentally about moving, transforming, and serving data. Position yourself as someone who understands data deeply rather than someone changing careers.

2. Learn the modern engineering stack

Beyond DP-700, I'd prioritize:

• Python (data processing, APIs, automation)
• PySpark/Spark (distributed data processing)
• Git and CI/CD
• Lakehouse concepts (Delta Lake, Parquet, partitioning)
• Data modeling (star schema, dimensional modeling, medallion architecture)
• Data orchestration (Fabric Data Factory, Airflow, or ADF)
• Data quality, monitoring, and governance
• Basic DevOps practices (source control, deployment pipelines)

  If you're targeting Microsoft Fabric roles specifically, become comfortable with:

  • Lakehouse
  • Notebooks
  • Spark
  • Data Factory
  • Warehouse
  • Eventstreams
  • OneLake
  • Security and governance
  • Semantic models

    3. Build projects that resemble real production work

    Instead of simple ETL demos, create projects that solve realistic business problems.

    Examples include:

    • End-to-end medallion architecture (Bronze → Silver → Gold)
    • Incremental ingestion with CDC
    • Batch and streaming pipeline together
    • Data quality framework with logging and error handling
    • Metadata-driven ingestion framework
    • CI/CD deployment using Git
    • Monitoring and alerting
    • Power BI semantic model built on the curated data

      These demonstrate engineering maturity far better than loading a CSV into a database.

      4. Reframe your resume

      Don't market yourself primarily as a DBA.

      Instead, highlight accomplishments such as:

      • Built scalable data pipelines
      • Optimized ETL performance
      • Designed enterprise data models
      • Automated database operations
      • Implemented security and governance
      • Improved data reliability and availability

        Hiring managers care more about outcomes than job titles.

        5. Demonstrate your work publicly

        Create a GitHub repository with:

        • Architecture diagrams
        • Source code
        • Documentation
        • Deployment steps
        • Sample datasets
        • Design decisions and trade-offs

          A well-documented portfolio often has more impact than another certification.

          Finally, remember that the market is competitive, but organizations still need experienced professionals who understand data. Your domain expertise is a significant advantage. The goal is to show that you can apply that experience using modern cloud platforms like Microsoft Fabric—not to convince employers you're starting from scratch.

          Best of luck with DP-700 and your transition!

          If this post helps, then please appreciate giving a Kudos or accepting as a Solution to help the other members find it more quickly.
          If I misunderstand your needs or you still have problems on it, please feel free to let me know. Thanks a lot!

@sandeeppri